System and method for automatic detection and localization of 3D bumps in medical images

Abstract

A method of detecting irregularities in a digitized image includes selecting a volume of interest in an image, growing a surface from a point in said volume of interest, determining the local gradient of each point on said surface, and applying a local filter to said gradient, determining a histogram of filtered gradients, and detecting an irregularity by selecting those points corresponding to gradient histogram values above a threshold value.

Description

CROSS REFERENCE TO RELATED UNITED STATES APPLICATIONS[0001]This application claims priority from “Automatic detection of 3D bumps for accurate localization in application such as lung nodules, colon polyps detection / segmentation”, U.S. Provisional Application No. 60 / 622,972 of Benjamin Odry, filed Oct. 28, 2004, the contents of which are incorporated herein by reference.TECHNICAL FIELD[0002]This invention is directed to the detection of irregularities in digitized medical images.DISCUSSION OF THE RELATED ART[0003]The diagnostically superior information available from data acquired from current imaging systems enables the detection of potential problems at earlier and more treatable stages. Given the vast quantity of detailed data acquirable from imaging systems, various algorithms must be developed to efficiently and accurately process image data. With the aid of computers, advances in image processing are generally performed on digital or digitized images.[0004]Digital images are c...

AI Technical Summary

Benefits of technology

The patent describes methods and systems for detecting abnormalities on a surface, such as bumps, using a digitized image. The method involves analyzing the image and identifying points with a 3D surface. The system can detect bumps along the pleura surface, which is the lung pleura, or detect polyp or aneurysm on vessel surfaces. The method uses a 3D volume of information and knowledge-based analysis to improve accuracy. The system can also detect irregularities in a histogram of gradients computed over a 3×3×3 neighborhood of each surface point. The method can be used as a first step for nodule segmentation and achieves an average segmentation consistency rate of 90% when tested on 20 nodules. The patent also describes a program storage device that can be used to perform the method.

Problems solved by technology

However, analyzing the datasets through 2D slices can be time consuming, and automated detection tools have been proven to be of help for physicians.

One challenge for segmenting lung nodules attached to the pleura is not to select the pleura voxels.

The inconveniences of 2D are that it is not always accurate and that the segmentation can vary from one slice to another.

Images